Answers fall into at least two categories, both with broader implications. First, a single poll question on a complex topic rarely reveals a complete portrait of public opinion. Second, issues rarely decide elections.
The apparent contradiction came into boldest relief on Super Tuesday when voters in exit polls supported Sanders’s marquee issue — “replacing all private health insurance with a single government plan for everyone” — by a 17-point margin.
The exit pollsters’ question offered a clear, up-or-down choice on the Sanders plan, but didn’t reflect the real debate. The former vice president was not simply opposed to Sanders’s approach, he offered his own, adding a public option to the Affordable Care Act, among other reforms.
A New York Times poll found 58 percent of Democrats preferred a government-run public plan alongside private plans — as per Biden — while just 25 percent wanted a Sanders-like, single government plan for all and 15 percent preferred no government plan at all.
It’s quite possible, for voters to support “replacing all private health insurance with a single government plan for everyone” over the implicit choice — doing nothing — while preferring a public option to a government-only plan.
However, the point at issue goes deeper than just poll questions.
Despite the fairy tale with which we are raised, while issues are the language we speak in campaigns, they’re rarely the decisive factor in elections.
A few years ago, we had the honor of helping pass an initiative requiring mandatory background checks for gun purchasers in Nevada.
Easy, you think. You’ve seen the national data. Some 90 percent of Americans favor the idea. Add a little formal initiative language, which always steals away at least a few votes, and support in Nevada still hovered around 80 percent.
Nonetheless, early on we warned our clients not to be deceived. Voters were happy to support the principle of universal background checks, but things could change dramatically if the initiative was transformed into a matter of identity.
Just over half the Nevada electorate were either gun owners or Republicans — and if those groups came to understand opposition to the initiative as an expression of these identities, their support for the principle would matter little.
That’s just what happened. What began as a 10-point deficit among Republicans exploded to nearly a 50-point margin for opponents.
The story was similar among gun-owning households who were initially tied but moved to overwhelming opposition.
Recognize that these shifts were not the result of changed attitudes toward the issue or the ballot measure itself.
At the outset of the campaign, 67 percent of Nevadans believed criminals would still be able to get guns if the measure passed, and in the end 68 percent held that view. Forty-six percent believed the measure would save lives at the beginning, as did 45 percent at the end. Thirty-five percent thought Second Amendment rights would be weakened at the outset and exactly the same number held that view election eve.
Though support for the background check initiative tumbled from 80 percent to a 50.4 percent victory on Election Day, the drop had little to do with the substance of the issue.
If votes on issues aren’t about the issues, how can we imagine that votes for candidates are? In fact, candidate elections are rarely about issues — voters often don’t know the candidates’ positions, don’t understand the issues and don’t care enough about most of them to let issues decide their votes.
More often voters cast their ballots based on identity (the candidates’ and the voters’), real world circumstances and emotional connection.
Neither Sanders nor Biden was likely to win “on the issues,” and neither did.
Mellman is President of The Mellman Group and has helped elect 30 U.S. senators, 12 governors and dozens of House members. Mellman served as pollster to Senate Democratic Leaders for over 20 years, as president of the American Association of Political Consultants, and is president of Democratic Majority for Israel.